Morphemes
Just as the phoneme is the smallest unit of sound in a language, the
“Isn’t that a word?” you may ask. “Isn’t a morpheme the same thing as a word?” Sometimes, but not always. Many words are monomorphemic, consisting of only one morpheme. But many are polymorphemic, consisting of more than one.
To see how this is possible, consider the word birdhouse. What is it made up of? You could think of it as a series of phonemes:
But no one thinks of the word birdhouse this way. Instead, you think of it like this:
A birdhouse, obviously, is a house for birds. When you were first learning about words in school, you probably learned to call this kind of word a compound word, and that’s accurate, since com- means ‘together’ and pound means ‘put’. A compound word, then, is a polymorphemic word. We make birdhouse by putting together bird and house. Each of those words is monomorphemic; their sum, birdhouse, is dimorphemic.
Can we split the word birdhouse any smaller than that? We could try chopping bird into two smaller parts,
Kinds of morphemes
We just saw that the word birdhouse, which has its own meaning (a house for birds), consists of two parts, each of which has its own meaning (a feathered animal, a building for living in). It’s easy to see that there are two morphemes in this word. But how many morphemes are in the word birdhouses? Two? Three?
The answer is three, and the reason is that, while the morpheme is indeed the smallest unit of meaning, there is more than one kind of meaning. That -s on the end has a meaning, and you can help me prove it. If I ask you what the -s does, you can answer, without even thinking, that it makes birdhouse plural – that is, it makes the word mean more than one house for birds, rather than the single structure meant by the word birdhouse. -s does not hold the same kind of meaning as bird and house do. It holds a less substantial, but equally necessary, meaning.
We can think of bird and house as “content” morphemes, since they embody concepts that we can picture in our heads, and of -s as a “function” morpheme, since it adds no extra image to the mental picture but merely adjusts the brushstrokes of the pictures that are already there. Linguists call content morphemes
Lexical meanings are, obviously, more important. When you were first learning your native language, the names you learned to call things were all lexical morphemes, and when you study a foreign language, the first words you learn are also lexical morphemes. That’s because while you can have lexical morphemes without grammatical morphemes, the reverse isn’t true: you can’t have grammatical without lexical. Lexical morphemes form the base of a word, which can then be tweaked in small ways by grammatical morphemes.
Soon we’ll look at the various ways that grammatical morphemes can tweak a word’s meaning, but first let’s familiarize ourselves with the structure of a word.
Word structure
At its most basic, a word consists of a
For example, birdhouse has a stem without an ending, while birdhouses has a stem (birdhouse) and an ending (-s). The ending is also called an
A stem doesn’t have to be a full word; it just has to be a lexical morpheme. Consider the word audacious, meaning ‘bold’:
The stem audac- is not a complete word in English. It nearly is in Latin – cap it with an -s to get audax, a Latin adjective that means ‘bold’. This word is itself a combination:
The base of this word is aud-. Though this piece is incomplete in Latin – you can’t just say “aud” and expect it to mean ‘bold’ in a sentence – it holds all the meaning of ‘bold’ in itself. -ac is an ending that creates an adjective stem from this base. We call it a
Here’s how audax’s descendant, audacious, breaks down in English:
Unlike Latin, English doesn’t use the lexical morpheme aud- as the base of a verb, so for the purpose of word formation, there’s no point for us in separating the stem audac- into its lexical and grammatical morphemes. That one piece, audac-, gives us a platform for building both the adjective audacious and its noun form, audacity:
We can say, for our purposes in English, that audacity contains two morphemes: audac-, which holds the meaning ‘bold’, and -ity, which turns that meaning into a noun. If we wanted to be especially thorough, we could then break audac- into two morphemes: the lexical aud-, which actually holds the meaning ‘bold’, and the grammatical -ac, which turns that meaning into an adjective. (It may seem strange to go to the trouble of making aud- into an adjective, only to make it into a noun! But that’s how this particular word evolved in Latin. It began as a verb base, was derived into an adjective stem, and only from there could be derived into a noun.)
Affixes
So far we’ve classified morphemes according to what kind of meaning they contribute to a word. If a morpheme gives us a core concept, something we can picture in our heads, it’s lexical; if it merely tweaks a concept that’s already there, it’s grammatical. At the same time, it’s also possible to classify morphemes on a different basis: whether they can stand alone as full words. If a morpheme can stand as a word, we call it a
On first thought, this way of classifying morphemes may seem unnecessary to you. Haven’t we said that you can’t have a grammatical morpheme without a lexical morpheme? If that’s true, aren’t all grammatical morphemes necessarily bound morphemes? Why bother pointing that out?
The concept of free versus bound is useful for two reasons. First, not all grammatical morphemes are bound. While it’s true that a grammatical morpheme in a vacuum is useless – that there must be some nearby lexical morpheme for it to work on – the lexical morpheme needn’t be in the same word. In the sentence I need to go to the store, though the second to could arguably be both lexical and grammatical, the and the first to are unquestionably grammatical. Neither is in the same word with the lexical morpheme it acts upon. The first to is a derivational morpheme that turns the verb go into a noun. The points out which particular store the speaker has in mind, but adds no extra concept to the one we already have of a place where goods are sold. Those two morphemes are free grammatical morphemes.
The second reason for the free/bound split is that not all lexical morphemes are free. A quick glance at audacious above is enough to prove that. Audac- (or, if you want to press the point even further, just aud-) is lexical in that it holds the full meaning of ‘bold’, but it can’t stand on its own as a word. For that, it needs an ending, either -ious or -ity.
Now that we have the concept of free and bound morphemes, we can give a precise definition to another useful concept: the affix. Simply put, an
There are four kinds of affixes, based on where they occur in a word:
Prefixes
Prefixes, found at the beginning of a word, usually add some kind of modifying meaning to the word’s lexical base. If that base is a noun, the prefix will hold an adjective-like meaning; if the base is a verb, the prefix will hold an adverb-like meaning.
Suffixes
Suffixes, found at the end of a word, can be inflectional or derivational in nature. Inflectional suffixes tweak a word’s meaning by customizing features like case, number, and gender for nouns and number, tense, and mood for verbs:
Derivational suffixes change a word’s part of speech to create a new word:
Infixes
Pretty much everyone who’s had any schooling in English knows about prefixes and suffixes, but did you know that an affix can go inside a word? About the only way we use infixes in English is to spice a word with profanity:
Circumfixes
An affix can also wrap around a word, working on it from both sides. This is rare in English, but it does happen:
Despite their scarcity in our language, circumfixes make intuitive sense to us. That’s how a playful coinage like embiggen, found neither in dictionaries nor in the mouths of real-life speakers, can be instantly understood.
Vowel mutation
Adding a piece to a lexical root is an obvious way to inflect a word, but another way, common in Germanic languages like English, is to change the root itself – specifically, the vowel. There are two kinds of vowel mutation,
Umlaut
When English was much older, its word for ‘foot’ was fōt. (Not surprising – we’ve just rewritten the long vowel ō as oo.) Some of the forms of this word put the vowel i at the end: *fōti-. (The asterisk means that we have no written record of this word but can tell from linguistic evidence that it existed.) Over time, speakers came to anticipate the vowel in the second syllable by pronouncing the vowel in the first syllable a little more like it, so as to reduce the effort of moving the tongue from one vowel position to another. (As we’ll see again and again, one of the greatest motivators of language change is laziness.) In other words, they started pronouncing the
What would this do to
The word umlaut suggests the process of scooting a vowel sound (Laut) over into the vicinity (um) of another vowel sound.
Ablaut
Not all vowel mutations arose by accident. Some have always been part of the inflectional system. As an English-speaker, you know the difference between “regular” verbs like walk/walked/walked and “irregular” verbs like sing/sang/sung.
Historically, though, there’s nothing irregular about sing and the other vowel-mutating verbs. In Old English, changing the root vowel was the default way to inflect verbs. There were several ways to do so – the patterns we still see in blow/blew/blown and throw/threw/thrown, ring/rang/rung and swim/swam/swum, ride/rode/ridden and write/wrote/written. Verbs that inflected by following these vowel-mutation patterns were called
If -ed verbs were so few and weak, why are they now so predominantly regular? The answer is easy to see when you consider how many verbs have come into English since the language began. Thanks to huge influxes from French and other languages, English now contains thousands more verbs than it used to. Whenever a new verb was added, it was inflected not according to one of the vowel-mutation schemes, which it wasn’t likely to fit, but simply by adding the suffix -ed. After enough words like this were added, the old weak verbs were so numerous as to be the new standard – what we call ”regular” verbs.
Meanwhile, many of the old strong verbs are now so rare that we’ve forgotten how to inflect them. Smite comes to mind. This archaic word for ‘strike’, preserved from extinction by frequent appearances in Arthurian romances and the King James Version of the Bible, is nonetheless so obscure today that whenever some enterprising copywriter tries to sound “medieval” by using it, he’s bound to inflect it incorrectly as a weak verb:
Interestingly, the past participle smitten has survived in a highly specialized sense: as an adjective meaning ‘head over heels in love’. When this is pointed out to you, it isn’t hard to see the connection between becoming infatuated and being physically struck. The well-preserved (since Roman times!) image of Cupid and his arrows reinforces this connection. But because it hardly ever gets pointed out, most people aren’t aware of it, and never do realize that smitten, a word not yet extinct but by no means still common, has anything to do with smite, a word they have one or two dim memories of at best. (And even if they did know, they’d probably still use the wrong participle, since “prepare to be smitten” sounds like someone’s about to seduce you!)
Typology
Languages are said to have morphological “types” based on how their morphemes are arranged. Which type a language belongs to depends on its morpheme-to-word ratio. A low MtW ratio means that the average word in the language has few morphemes, while a high MtW ratio means that the average word has many morphemes. Put another way, low MtW ratios mean few inflections and high MtW ratios mean many inflections. In order from low to high, the types are
Analytic
Analytic languages, sometimes called isolating languages, shy away from inflectional endings, preferring to color the meanings of words by adding other words rather than affixing grammatical morphemes.
The best example of an analytic language is Chinese, which doesn’t use inflections at all. This means, for example, that nouns don’t have separate forms for singular and plural number. (Many languages don’t make a big deal about number the way English does.) If a speaker needs to point out that there’s more than one of a thing, he adds another word consisting of a numeral and a counter. Similarly, Chinese verbs aren’t marked for features like tense. If a speaker needs to point out that an action happened in the past, he adds a time adverb like ‘yesterday’ or ‘last month’.
Though not (yet) entirely free of inflections the way Chinese is, Modern English is also an analytic language. A noun like fox has only four forms – fox, foxes, fox’s, and foxes’, and three of them sound the same. A verb like sing also has only four forms: sing, sings, singing, and sung. And a verb like walk has only three: walk, walks, and walking. If you’ve ever studied a European language like Spanish or an East Asian language like Japanese, you know that other languages can have far more word-forms than English does.
Synthetic
Compared with analytic languages, synthetic languages color the meanings of words by adding a chain of inflectional endings. There are two kinds of synthetic languages according to how distinct a word’s morphemes remain from each other. If the morphemes can be easily distinguished, preserving recognizable boundaries, the language is
Polysynthetic
In Inuktitut, the language of one of the indigenous peoples of Canada, tavvakiqutiqarpiit means ‘do you have any tobacco for sale?’. This whole-sentence-in-a-single-word phenomenon, scarcely believable to speakers of an analytic language, is the mark of a polysynthetic language. Morphemes line up in prescribed patterns, settling into place with context-sensitive shapes.
The natural question to ask, on seeing a sample of a polysynthetic language, is: How can we, not being Inuits, know that tavvakiqutiqarpiit is all one word, and not just a sentence without spaces? The answer would become a bit complicated if explained fully, but in brief it’s this: whereas an independent word keeps the same shape from sentence to sentence, even as it interacts with other words, the morphemes in a polysynthetic word-phrase can take different shapes based on how they interact with other morphemes in the same word-phrase.
